Re: The future of Spark . Spark 2014 : a wreckage

["Randy Brukardt" <randy@rrsoftware.com>]

"Robert A Duff" <bobduff@shell01.TheWorld.com> wrote in message 
news:wcck3kvdfgg.fsf@shell01.TheWorld.com...
> "Randy Brukardt" <randy@rrsoftware.com> writes:
>
>> I'm pretty amazed to find out that I can't debug concurrent programs,
>> because I've been doing it for decades.
>
> Are you done yet?

I presume this is supposed to be a joke of some kind (but without a 
emoticon, can't be sure), and I'm not sure what the joke is supposed to be, 
either.

So, the serious answer: The servers appear to work properly, without 
deadlocks or other problems. There can't be 100% certainty of that, but 
hardly anyone requires that sort of certainty. I'd like a form of 
correctness-by-construction for concurrent programs, but that appears beyond 
the state of the art. I don't see proof tools adding much in this area, the 
main problem is ensuring that bad things are avoided (about the only place 
that proof might help would be in preventing livelock, but that's way down 
the list of problems).

                       Randy.